Physician champions are critical to pulling off the installation of a computerized physician order-entry system, but theyre just as important to providing user support in the long years after going live, according to a joint report from the Massachusetts Technology Collaborative and the New England Healthcare Institute.
Then again, so are nurse informaticists. And, more and more, so are hospitalists.
Those planning a CPOE project also might want to consider getting some outside help in preparing and maintaining the computerized order sets and other clinical content, which, as it turns out, is a lot bigger deal going forward than they anticipated.
And definitely, while all six hospitals reported have electronic medication administration systems, all reported their medication reconciliation functions need considerable work before their IT systems will perform the reconciliations mandated by the Joint Commission with fluidity.
These and others are the lessons learned by six hospitalslarge and quite smallthat were early adopters of the still relatively rare technology and were part of a research study conduced by Computer Sciences Corp. on behalf of the two IT booster organizations.
According to the report, survey-participating hospitals, bed sizes, CPOE vendors, system rollout years and percentages of orders entered by physicians are:
- Newport (R.I.) Hospital, 129 beds, Siemens, 2004, 92%;
- Baystate Franklin Medical Center, Greenfield, Mass., 130 beds, Cerner Corp., 2004, 99%;
- St. Marys Health Care, Grand Rapids, Mich., 230 beds, Cerner Corp., 2004, 98%;
- Glens Falls (N.Y.) Hospital, 410 beds, Cerner Corp., 2006, 80%;
- Citizens Memorial Hospital, Bolivar, Mo., 74 beds, Meditech, 2003, 100%; and
- Alamance Regional Medical Center, Burlington, N.C., 238 beds, Eclipsys Corp., 2000, 87%.
The hope, according to the collaborative and the institute, is that the 21-page report will serve as a practical tool for use by CPOE project managers, CIOs, chief medical information officers and others responsible for the CPOE effort in their hospitals.
The report said there is no correlation between hospital size and the number of support people for CPOE. Ongoing, CPOE support staffs at the six hospitals varied from 2.5 full-time equivalents to eight more FTEs, whose members included the chief medical information officer or a physician of similar title or duties, or a physician who was named a CPOE leader. Nurse informaticist was the most common title among these CPOE support group members. But, during the initial rollout of CPOE, the hospitals found it critical to provide physicians with help at the elbow and to offer one-on-one coaching rather than classroom training. This same hands-on approach is applied now when major changes are made to CPOE and when physicians unfamiliar with CPOE are brought on board.
Tellingly, None of the study hospitals has had success with Web-based training for physicians because, even when it was available, few of the doctors would use it.
Since all of the CPOE systems in the study were from vendors of enterprisewide clinical IT systems, the CPOE systems were subject to one of the common drawbacks of such enterprise systemsthey work better in some departments than in others. Not surprisingly, then that, Teams from several hospitals noted that CPOE does not yet work smoothly in clinical areas such as ambulatory surgery or obstetrics because the workflow differs significantly from that of medical and surgical inpatient units. One of the study hospitals planned to contract with the CIS vendor for a care-optimization assessment that would produce recommendations for fine-tuning both care plans and workflows.
Only one of the six hospitals used an outside vendorZynxHealthto provide CPOE order sets. Zynx also provides an order set management tool. However, the CMIO from another hospital reported to the survey team that he regretted not getting outside help, adding, We vastly underestimate the effort that would be required.